Infants with epileptic seizures may have dramatic elevations of prolactin or beta endorphin levels compared to controls. Using these levels as indicators, diagnostic accuracy of neonatal seizures may be improved. This may be of particular interest in infants therapeutically paralyzed (with neuromuscular blocking agents) or in those infants who have a variety of paroxysmal attacks which are not clearly based on an epileptic mechanism. This study will test whether deviations of prolactin and beta endorphin occur following epileptic seizure.